Writing
Making AI Make Sense
A framework for deploying AI where it’s strong and routing around where it’s weak. These pieces build on each other.
Start here: Making AI Make Sense — The complete framework in one page.
Or explore the components:
| Essay | What it covers |
|---|---|
| Knowledge as Capability | The foundational claim: knowledge is capability to produce outcomes |
| Agent-Relative Tacitness | Why tacitness depends on who’s doing the knowing |
| Tacit Space Shrinkage | How AI is making the inexpressible expressible |
| AI as Oracle vs. Assistant | The fundamental choice: ask for answers or ask for execution |
| The Strong Oracle Trap | Why sophisticated dialogue doesn’t solve the verification problem |
| Executable Knowledge Architecture | The pattern for AI-augmented professional work |
| Capability Governance | The governance framework: stakeholder obligations for generated artifacts |
| Agentic AI as Universal Interface | How the pattern scales: AI removes barriers to direct system access |
| The Universal Interface Thesis | Beyond knowledge work: the pattern applies to any control surface |
| Automating Expertise Gets Easier and Easier | What faces pressure and what remains as more expertise becomes software |
| Ontology Generation | How organizational knowledge accumulates from universal interface logs |
| What Benchmarks Aren’t Measuring | Why current AI benchmarks test the wrong things for professional work |
| AI-First Software | The design philosophy: AI as foundation, not feature |
| The Return of the Assistant | The DIY knowledge worker made sense. AI offers something better. |
| Update for February 2026 | Two years of advances tested the framework. The distinction holds—but agentic AI creates new gaps. |
| Answers to Critics | Direct responses to adversarial critique: what’s refined, what’s defended, what’s unchanged. |
Updates and Responses
The framework has been tested against critique and technological change. These essays document the evolution.
| Essay | What it covers |
|---|---|
| Update for February 2026 | Research validating the oracle/assistant distinction against 2024-2026 advances; the agentic erosion problem |
| Answers to Critics | Direct engagement with adversarial review: refinements, precision improvements, and defended positions |
Standalone Essays
Other writing, not part of the framework above.
| Essay | Topic |
|---|---|
| Parrots Are All You Need | Why “stochastic parrot” isn’t the insult people think it is |
| LLM Stochasticity and Determinism | Understanding the randomness in language models |
| The Consulting Threat and Opportunity | What AI means for professional services |
| Critical Thinking Rules | Principles for clear reasoning |
| The Support Team’s Real Job | What support work is actually about |
| Leadership Lessons from Science Fiction | What SF teaches about leading |
| The Remote Work Formula | Making distributed work work |
| AI Table Extraction Comparison | Testing AI on a specific task |
| The Office Availability Math | Why 80% in the office means 40% available |
Defining concepts and core distinctions
- Knowledge as Capability: An Operational Foundation for Human-AI Cooperation Defining knowledge as capability to produce outcomes, the foundational claim that makes human-AI cooperation coherent.
- AI as Oracle vs. Assistant: Two Patterns of Deployment The fundamental choice in how you use AI determines whether you can govern it. Oracle pattern creates verification problems. Assistant pattern solves them.
- Parrots Are All You Need Why generative AI will introduce greater change, faster, than the Internet itself—and what that means for knowledge workers.
The central frameworks
- The Universal Interface Thesis Code generation isn't just for knowledge work. It's the architecture for human interaction with any control surface—shop floors, building systems, medical devices, enterprise software. The implications restructure how we think about training, expertise, and vendor lock-in.
- The Return of the Assistant The DIY knowledge worker made sense for forty years. AI assistants offer something better: the freedom of DIY with the leverage of delegation.
- Ontology Generation When the AI tool is the universal interface, comprehensive logs reveal both what the organization does and what concepts it uses. Structure and procedure emerge together.
- Automating Expertise Gets Easier and Easier As agentic AI lowers the barriers, more expertise becomes software. What faces pressure, what remains, and where value moves.
- Agentic AI as Universal Interface Software is distilled expertise. Two barriers have prevented most expertise from becoming software: coding skill and system access knowledge. Agentic AI removes both.
- Executable Knowledge Architecture A framework for integrating AI into professional practice that preserves expertise, ensures reproducibility, and closes the governance gap between AI capabilities and organizational needs.
- What Benchmarks Aren't Measuring GDPval and similar benchmarks test well-specified problems with verifiable answers. They don't—can't—test the ambiguous, contested, judgment-intensive work where professional expertise commands premiums. The measurement gap is structural, not temporary.
- Capability Governance: A Framework for Generated Artifacts A framework for governing generated artifacts through defensibility. Two routes: exposed methodology or empirical track record. Obligations for producers, consumers, underwriters, and regulators.
- The Strong Oracle Trap Thorough AI dialogue feels like it solves the verification problem. It doesn't. Understanding why requires seeing oracle pattern as a spectrum, and recognizing what tools can and cannot fix.
Extending the foundations
- Tacit Space Shrinkage: When AI Articulates the Inexpressible As AI capabilities expand, the domain of knowledge that organizations cannot express or operationalize is shrinking, with real implications for expertise and governance.
- Agent-Relative Tacitness: A New Framework Rethinking what makes knowledge tacit or explicit, and why the answer depends on who is doing the knowing.
Evidence and experiments
- The Office Availability Math Two people in the same office are available to each other far less than you'd think. The math explains why DIY tools were liberating.
- Stochastic in Form, Deterministic in Function 700 iterations across seven models. 25% accuracy asking directly. 100% accuracy asking for code. The gap tells you everything about how to use these tools.
- Evaluating AI Table Extraction: Claude vs Docling vs ScaleDP A practical comparison of three table extraction approaches reveals surprising accuracy gaps that practitioners need to understand before selecting tools for document processing at scale.
Applying the framework
- AI-First Software: Engineering the Post-Interface Era Generative AI has solved language. That's enough to change software design—not in the future, but right now.
- Existential Threat to Consulting = Huge Opportunity How AI-encoded expertise disrupts consulting's traditional basis for premium fees, and what firms need to do about it.
Lessons from experience
- Mike's Rules for Critical Thinking A working set of rules for critical thinking organized into three domains: foundation, analysis, and engagement. Written for non-academics who want a down-to-earth framework.
- Leadership Lessons from Science Fiction The books that most shaped how I lead weren't management classics—they were science fiction novels I read in my youth.
- The Remote Work Formula Individual flexibility and autonomy matter. But so does group consideration. The formula isn't complicated.
- The Real Job of Support Teams in Professional Services Support teams exist to protect the scarcest resource in professional services: the time professionals spend not working.
- EKA in Practice The framework is no longer theoretical. Here's what we learned when we deployed it.
Framework validation and evolution
Engaging critique and refining claims